Media content recommendation method and device, equipment and storage medium

By calculating the characteristic vector similarity value of the target media content and candidate media content, and recommending it based on multimodal information and user data, the problem that users' diversified needs in the prior art is difficult to meet and the user experience is improved.

CN118916501AInactive Publication Date: 2024-11-08BEIJING VOLCANO ENGINE TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411036204.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-30
Publication Date
2024-11-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing technology is difficult to meet the diverse content recommendation needs of users and lacks rich ways to recommend media content.

Method used

By receiving the media content recommendation request, the feature vector similarity value of the target media content and the candidate media content is calculated, and the recommended media content is determined based on the similarity value, and multimodal information and user data are recommended.

Benefits of technology

It enriches the recommendation methods of media content, meets users' diverse content recommendation needs, and improves user experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916501A_ABST
    Figure CN118916501A_ABST
Patent Text Reader

Abstract

The invention provides a media content recommendation method and device, equipment and a storage medium, and the method comprises the steps: receiving a media content recommendation request, the media content recommendation request carries an identifier of target media content, and the target media content is determined in response to a triggering operation for a target notification message; and determining a similarity value between the feature vector of the target media content and the feature vector of the candidate media content, determining recommended media content based on the similarity value corresponding to the candidate media content, and returning the recommended media content in response to the media content recommendation request. Visibly, according to the embodiment of the invention, after the target media content is viewed through the target notification message and enters the media content recommendation scene, the recommended media content is determined based on the similarity between the target media content and the candidate media content, the media content recommendation modes are enriched, diversified content recommendation requirements of users are met, and the user experience is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the field of data processing, and in particular to a media content recommendation method, device, equipment and storage medium. Background Art

[0002] With the continuous development of computer technology, there are more and more content recommendation applications, and users' demands for content recommendation are becoming more and more diverse.

[0003] Therefore, how to further enrich the implementation methods of content recommendation to meet users' diverse content recommendation needs has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] In order to solve the above technical problems, the embodiments of the present disclosure provide a media content recommendation method, apparatus, device and storage medium.

[0005] In a first aspect, the present disclosure provides a media content recommendation method, the method comprising:

[0006] Receive a media content recommendation request; wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content;

[0007] Calculating a similarity value between a feature vector of the target media content and a feature vector of a candidate media content, and determining a recommended media content based on the similarity value corresponding to the candidate media content;

[0008] In response to the media content recommendation request, the recommended media content is returned.

[0009] In an optional implementation manner, before calculating the similarity value between the feature vector of the target media content and the feature vector of the candidate media content, the method further includes:

[0010] Determine a feature vector of the target media content based on the multimodal information of the target media content; wherein the multimodal information of the target media content includes information of multiple target dimensions extracted from the target media content;

[0011] And, based on the multimodal information of the candidate media content, determining the feature vector of the candidate media content; wherein the multimodal information of the candidate media content includes information of the multiple target dimensions extracted from the candidate media content.

[0012] In an optional implementation manner, before calculating the similarity value between the feature vector of the target media content and the feature vector of the candidate media content, the method further includes:

[0013] Based on the identification of the target media content, a feature vector of the target media content is obtained from a target database; wherein the feature vector of the target media content is determined in advance based on the multimodal information of the target media content, and the target database also stores a feature vector determined based on the multimodal information of the candidate media content;

[0014] Accordingly, the calculating of the similarity value between the feature vector of the target media content and the feature vector of the candidate media content includes:

[0015] The similarity values ​​between the feature vectors of the target media content and the feature vectors of the candidate media content in the target database are calculated.

[0016] In an optional implementation, the multimodal information includes at least two of the following: a media content title, a media content text, a media content picture, and a media content video frame.

[0017] In an optional implementation manner, determining the recommended media content based on the similarity values ​​corresponding to the candidate media content includes:

[0018] Adding the candidate media content whose similarity value is not less than a preset first threshold value into a target data pool; wherein the target data pool at least includes candidate media content determined based on target user data, and the target user data includes user interaction data and / or user interest tags;

[0019] Calculating similarity values ​​between the feature vectors of the target media content and the feature vectors of the candidate media content in the target data pool;

[0020] The candidate media content in the target data pool whose similarity value is not less than a preset second threshold is determined as the recommended media content.

[0021] In an optional implementation manner, before determining the candidate media content in the target data pool whose similarity value is not less than a preset second threshold as the recommended media content, the method further includes:

[0022] Determine the matching degree values ​​between the target user data and the candidate media content in the target data pool respectively;

[0023] Correspondingly, determining the candidate media content in the target data pool whose similarity value is not less than a preset second threshold as the recommended media content includes:

[0024] Determining a fusion score value corresponding to the candidate media content based on the similarity value and the matching degree value corresponding to the candidate media content in the target data pool;

[0025] The candidate media content in the target data pool whose fusion score value is not less than a preset third threshold is determined as the recommended media content.

[0026] In an optional implementation manner, the candidate media content belongs to any of the following content themes: article theme, video theme, and graphic theme.

[0027] In a second aspect, the present disclosure further provides a media content recommendation device, the device comprising:

[0028] A receiving module, configured to receive a media content recommendation request; wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content;

[0029] A first determination module, configured to calculate a similarity value between a feature vector of the target media content and a feature vector of a candidate media content, and determine a recommended media content based on the similarity value corresponding to the candidate media content;

[0030] The returning module is configured to return the recommended media content in response to the media content recommendation request.

[0031] In a third aspect, the present disclosure provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device implements the above method.

[0032] In a fourth aspect, the present disclosure provides a media content recommendation device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above method when executing the computer program.

[0033] In a fifth aspect, the present disclosure provides a computer program product, wherein the computer program product comprises a computer program / instructions, and the computer program / instructions implement the above method when executed by a processor.

[0034] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has at least the following advantages:

[0035] The disclosed embodiments provide a method for recommending media content. Specifically, first, a media content recommendation request is received, wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content. Then, a similarity value between a feature vector of the target media content and a feature vector of a candidate media content is calculated, and recommended media content is determined based on the similarity value corresponding to the candidate media content. Then, in response to the media content recommendation request, the recommended media content is returned.

[0036] It can be seen that after the target media content is viewed through the target notification message to enter the media content recommendation scenario, the embodiment of the present disclosure determines the recommended media content based on the similarity between the target media content and the candidate media content, enriching the media content recommendation method, meeting the user's diverse content recommendation needs, and improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0039] Figure 1 A processing flow chart of a media content recommendation method provided in some embodiments of the present disclosure;

[0040] Figure 2 A schematic diagram of a process for obtaining a feature vector of target media content provided by some embodiments of the present disclosure;

[0041] Figure 3 A schematic diagram of a process for obtaining a feature vector of candidate media content provided by some embodiments of the present disclosure;

[0042] Figure 4 A data interaction diagram of a media content recommendation method provided by some embodiments of the present disclosure;

[0043] Figure 5 A schematic diagram of the structure of a media content recommendation device provided in some embodiments of the present disclosure;

[0044] Figure 6 A schematic diagram of the structure of a media content recommendation device provided in some embodiments of the present disclosure. DETAILED DESCRIPTION

[0045] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0046] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0047] In related technologies, update notifications of media content (such as news, information, etc.) can be sent to the client via notification messages. When the client receives a trigger operation from the user for the notification message, it not only displays the media content corresponding to the notification message, but also recommends media content that can be continued to be browsed by the user.

[0048] To this end, an embodiment of the present disclosure provides a media content recommendation method, first, receiving a media content recommendation request, wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content, then, calculating a similarity value between a feature vector of the target media content and a feature vector of a candidate media content, and determining recommended media content based on the similarity value corresponding to the candidate media content, and then returning the recommended media content in response to the media content recommendation request.

[0049] It can be seen that after the target media content is viewed through the target notification message to enter the media content recommendation scenario, the embodiment of the present disclosure determines the recommended media content based on the similarity between the target media content and the candidate media content, enriching the media content recommendation method, meeting the user's diverse content recommendation needs, and improving the user experience.

[0050] Specifically, the present disclosure provides a media content recommendation method, which can be executed by a media content recommendation device, wherein the device can be implemented by software and / or hardware deployed on a server. Figure 1 FIG. 1 is a flowchart of a media content recommendation method provided by an embodiment of the present disclosure, which specifically includes:

[0051] S101: receiving a media content recommendation request.

[0052] The media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation on a target notification message, and the target notification message is used to notify the target media content.

[0053] The media content recommendation request in the embodiment of the present disclosure is used to request the recommendation of media content. The media content recommendation request carries the identifier of the target media content. Based on the identifier of the target media content, the target media content can be determined, and the target media content is used to recommend the media content.

[0054] The target media content is determined based on the triggering operation on the target notification message. The target notification message is the notification message currently clicked by the user, and the notification message is used to prompt the user to view the corresponding media content, that is, by triggering the notification message, the media content corresponding to the notification message can be viewed.

[0055] S102: Determine a similarity value between a feature vector of the target media content and a feature vector of candidate media content, and determine recommended media content based on the similarity value corresponding to the candidate media content.

[0056] The feature vector of the target media content in the embodiment of the present disclosure is used to characterize the content features of the target media content.

[0057] In an optional implementation, the feature vector of the target media content may be determined based on multimodal information of the target media content, wherein the multimodal information of the target media content may include information of multiple dimensions extracted from the target media content.

[0058] In an optional implementation, the multimodal information may include at least two of the following: a media content title, a media content text, a media content picture, and a media content video frame.

[0059] In practical applications, in order to calculate the similarity value between the feature vector of the target media content and the feature vector of the candidate media content, the embodiment of the present disclosure also needs to obtain the feature vector of the target media content.

[0060] In an optional implementation, after the multimodal information is extracted from the target media content, the multimodal information is input into the multimodal model, and after being processed by the multimodal model, a feature vector of the target media content is output. The multimodal model is obtained through multimodal sample training, and the specific model training process is not described in detail in the embodiment of the present disclosure. Figure 2 FIG. 1 is a schematic diagram of a process for obtaining a feature vector of target media content provided by an embodiment of the present disclosure.

[0061] In an optional implementation, for a method of acquiring target media content, upon receiving a media content recommendation request, the target media content may be determined from a media content database based on an identifier of the target media content carried in the media content recommendation request, wherein the media content database is used to store media content. Multimodal information of the target media content may be extracted, and a feature vector of the target media content may be calculated based on the multimodal information.

[0062] In addition, the embodiment of the present disclosure also needs to obtain a feature vector of the candidate media content for calculating a similarity value with the feature vector of the target media content.

[0063] The candidate media content may be media content stored in a media content database, and the candidate media content may be media content belonging to any content theme. Specifically, the candidate media content may be media content of article theme, video theme, or graphic theme.

[0064] The feature vector of the candidate media content is used to reflect the content features of the candidate media content. The feature vector is determined based on multimodal information of the candidate media content. The multimodal information may include information of multiple dimensions extracted from the candidate media content.

[0065] In an optional implementation, feature vectors of candidate media content may be pre-stored in a feature vector library, which stores feature vectors of each media content in the above media content database. Each media content in the media content database may serve as candidate media content.

[0066] like Figure 3 The figure is a schematic diagram of a process of obtaining a feature vector of a candidate media content provided by an embodiment of the present disclosure. Specifically, for each candidate media content in the media content database, after extracting the multimodal information of the candidate media content from the candidate media content, the multimodal information is input into the multimodal model, and the feature vector of the candidate media content is obtained after being processed by the multimodal model, and the feature vector is stored in the feature vector library.

[0067] After obtaining the feature vector of the target media content and the feature vector of the candidate media content, a similarity value between the feature vector of the target media content and the feature vector of the candidate media content is calculated.

[0068] The similarity value is used to characterize the similarity between the target media content and the candidate media content. The method for calculating the similarity value based on the feature vector of the target media content and the feature vector of the candidate media content is not limited in any way in the embodiments of the present disclosure.

[0069] In an optional implementation, after calculating the similarity value between the feature vector of the target media content and the feature vector of the candidate media content, the candidate media content can be sorted according to the similarity value, and the candidate recommended content with a high similarity value ranking can be returned to the client as the recommended media content of the target media content, for example, the top N candidate recommended content can be used as the recommended media content of the target media content.

[0070] In another optional implementation, a portion of candidate media content with a higher similarity value can be screened out through the similarity value of the feature vector, and the portion of candidate media content can be added to the target data pool. Specifically, candidate media content with a similarity value not less than a preset first threshold can be added to the target data pool. The target data pool also includes candidate media content determined based on other methods, and the candidate media content in the target data pool is used to determine the recommended media content of the target media content.

[0071] Among them, determining the candidate media content based on other methods may include determining the candidate media content based on target user data, and the target user data may include user interaction data, user interest tags, etc.

[0072] The preset first threshold is a predetermined value used to screen out candidate media content having a high similarity with the feature vector of the target media content.

[0073] Then, for each candidate media content in the target data pool, the similarity values ​​between the feature vectors of the target media content and the feature vectors of the candidate media content in the target data pool are calculated, and then, the candidate media content with a similarity value not less than a preset second threshold is screened out from the candidate media content in the target data pool, and the recommended media content of the target media content is determined from the screened candidate media content and returned to the client. The preset second threshold and the preset first threshold may be the same or different.

[0074] In an optional implementation, in addition to calculating the similarity value between the feature vector of the target media content and the candidate media content in the target data pool, the matching degree value between the candidate media content in the target data pool and the target user data can also be calculated based on the target user data, and then based on the similarity value and the matching degree value, the recommended media content can be screened from the candidate media content in the target data pool. The matching degree value between the target user data and the candidate media content is used to represent the user's interest in the candidate media content.

[0075] Optionally, the target user data and the candidate media content in the target data pool are input into the model, and the matching degree value between the target user data and the candidate media content in the target data pool is output after being processed by the model.

[0076] Then, based on the similarity values ​​and matching degree values ​​corresponding to the candidate media content in the target data pool, the fusion score value corresponding to the candidate media content is calculated, and based on the fusion score value corresponding to the candidate media content, the candidate media content whose fusion score value is not less than a preset third threshold is determined as the recommended media content.

[0077] The preset third threshold, the preset second threshold and the preset first threshold may be the same or different.

[0078] In an optional implementation, weights may be set for the similarity values ​​and matching values ​​corresponding to the candidate media contents, respectively, and the fusion score value may be calculated based on the similarity values ​​and matching values ​​corresponding to the candidate media contents and the corresponding weights.

[0079] In actual applications, since the multimodal information of media content can reflect the similarity between media content of different themes, the method of calculating the feature vectors of target media content and candidate media content based on the multimodal information of media content can enable the determined recommended media content to be recommended across themes. Therefore, the recommended media content determined in the embodiments of the present disclosure may include media content in the form of different content themes. Specifically, the recommended media content may include article themes, video themes, graphic themes, audio themes, etc.

[0080] S103: In response to the media content recommendation request, return the recommended media content.

[0081] In the embodiment of the present disclosure, after receiving a media content recommendation request from a client, recommended media content of the determined target media content is returned to the client.

[0082] In addition, when a media content recommendation request is triggered by a target notification message, the target media content corresponding to the target notification message needs to be returned to the client to display the target media content corresponding to the target notification message to the user. On the basis of displaying the target media content, browsing of the above-mentioned recommended media content is also supported.

[0083] In the media content recommendation method provided by the embodiments of the present disclosure, specifically, a media content recommendation request is received, wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content, then, a similarity value between a feature vector of the target media content and a feature vector of a candidate media content is calculated, and recommended media content is determined based on the similarity value corresponding to the candidate media content, and then, in response to the media content recommendation request, the recommended media content is returned.

[0084] It can be seen that after the target media content is viewed through the target notification message to enter the media content recommendation scenario, the embodiment of the present disclosure determines the recommended media content based on the similarity between the target media content and the candidate media content, enriching the media content recommendation method, meeting the user's diverse content recommendation needs, and improving the user experience.

[0085] In order to facilitate understanding of the above-mentioned embodiment, a data interaction diagram of a media content recommendation method provided by the present disclosure is shown in FIG. Figure 4 First, a notification message is displayed on the client on the current user device. When the client receives a trigger operation for the target notification message, the client sends a media content recommendation request to the server, and the media content recommendation request carries an identifier of the target media content;

[0086] When the server receives a media content recommendation request, it obtains the target media content corresponding to the target notification message based on the media content identifier carried in the media content recommendation request, and the server determines the recommended media content based on the target media content. Then, the server returns the target media content corresponding to the target notification message and the recommended media content to the client.

[0087] When the client receives the target media content and the recommended media content corresponding to the target notification message, they are displayed on the display interface of the client.

[0088] In one application scenario, the target media content corresponding to the target notification message and the recommended media content can form a media content stream, and the target media content corresponding to the target notification message is the first media content in the media content stream. When a sliding operation is received for the target media content, the recommended media content is displayed.

[0089] In another application scenario, when the target media content corresponding to the target notification message is displayed on the display interface of the client, the recommended media content of the target media content is displayed in a preset area of ​​the display interface in the form of a recommendation list for user browsing.

[0090] Based on the above method embodiment, the present disclosure also provides a media content recommendation device, referring to Figure 5 , is a schematic diagram of the structure of a media content recommendation device provided by an embodiment of the present disclosure, the device comprising:

[0091] The receiving module 501 is used to receive a media content recommendation request; wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content;

[0092] A first determination module 502, configured to calculate a similarity value between a feature vector of the target media content and a feature vector of a candidate media content, and determine a recommended media content based on the similarity value corresponding to the candidate media content;

[0093] The returning module is configured to return the recommended media content in response to the media content recommendation request.

[0094] In an optional implementation, the device further includes:

[0095] A second determination module, configured to determine a feature vector of the target media content based on the multimodal information of the target media content; wherein the multimodal information of the target media content includes information of multiple target dimensions extracted from the target media content;

[0096] The third determination module is used to determine the feature vector of the candidate media content based on the multimodal information of the candidate media content; wherein the multimodal information of the candidate media content includes information of multiple target dimensions extracted from the candidate media content.

[0097] In an optional implementation, the device further includes:

[0098] an acquisition module, configured to acquire a feature vector of the target media content from a target database based on an identifier of the target media content; wherein the feature vector of the target media content is determined in advance based on multimodal information of the target media content, and the target database further stores a feature vector determined based on multimodal information of candidate media content;

[0099] Accordingly, the first determining module is specifically configured to:

[0100] The similarity values ​​between the feature vectors of the target media content and the feature vectors of the candidate media content in the target database are calculated.

[0101] In an optional implementation, the multimodal information includes at least two of the following: a media content title, a media content text, a media content picture, and a media content video frame.

[0102] In an optional implementation manner, the first determining module includes:

[0103] An adding submodule, configured to add the candidate media content whose similarity value is not less than a preset first threshold value to a target data pool; wherein the target data pool at least includes candidate media content determined based on target user data, and the target user data includes user interaction data and / or user interest tags;

[0104] A calculation submodule, used for calculating similarity values ​​between the feature vectors of the target media content and the feature vectors of the candidate media content in the target data pool;

[0105] The first determination submodule is configured to determine the candidate media content in the target data pool whose similarity value is not less than a preset second threshold as the recommended media content.

[0106] In an optional implementation, the device further includes:

[0107] A fourth determination module, used to determine the matching degree values ​​between the target user data and the candidate media content in the target data pool;

[0108] Accordingly, the first determining submodule includes:

[0109] A second determination submodule, configured to determine a fusion score value corresponding to the candidate media content based on the similarity value and the matching degree value corresponding to the candidate media content in the target data pool;

[0110] The third determination submodule is configured to determine the candidate media content in the target data pool whose fusion score value is not less than a preset third threshold as the recommended media content.

[0111] In an optional implementation manner, the candidate media content belongs to any of the following content themes: article theme, video theme, and graphic theme.

[0112] In the media content recommendation device provided by the embodiment of the present disclosure, first, a media content recommendation request is received, wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content, then, a similarity value between a feature vector of the target media content and a feature vector of a candidate media content is calculated, and recommended media content is determined based on the similarity value corresponding to the candidate media content, and then the recommended media content is returned in response to the media content recommendation request.

[0113] It can be seen that after the target media content is viewed through the target notification message to enter the media content recommendation scenario, the embodiment of the present disclosure determines the recommended media content based on the similarity between the target media content and the candidate media content, enriching the media content recommendation method, meeting the user's diverse content recommendation needs, and improving the user experience.

[0114] In addition to the above-mentioned method and apparatus, the embodiment of the present disclosure further provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal device, the terminal device implements the media content recommendation method described in the embodiment of the present disclosure.

[0115] The embodiment of the present disclosure further provides a computer program product, which includes a computer program / instructions. When the computer program / instructions are executed by a processor, the media content recommendation method described in the embodiment of the present disclosure is implemented.

[0116] In addition, the present disclosure also provides a media content recommendation device, see Figure 6 As shown, it may include:

[0117] Processor 601, memory 602, input device 603 and output device 604. The number of processors 601 in the media content recommendation device can be one or more. Figure 6 In some embodiments of the present disclosure, the processor 601, the memory 602, the input device 603 and the output device 604 may be connected via a bus or other means, wherein: Figure 6 The example of connecting through bus is taken in the following.

[0118] The memory 602 can be used to store software programs and modules. The processor 601 executes various functional applications and data processing of the media content recommendation device by running the software programs and modules stored in the memory 602. The memory 602 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc. In addition, the memory 602 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. The input device 603 can be used to receive input digital or character information, and generate signal input related to user settings and function control of the media content recommendation device.

[0119] Specifically in this embodiment, the processor 601 will load the executable files corresponding to the processes of one or more applications into the memory 602 according to the following instructions, and the processor 601 will run the applications stored in the memory 602, thereby realizing various functions of the above-mentioned media content recommendation device.

[0120] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0121] The above description is only a specific embodiment of the present disclosure, so that those skilled in the art can understand or implement the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A media content recommendation method, characterized in that: include: Receive a media content recommendation request; wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content; Determining a similarity value between a feature vector of the target media content and a feature vector of a candidate media content, and determining recommended media content based on the similarity value corresponding to the candidate media content; In response to the media content recommendation request, the recommended media content is returned.

2. The method according to claim 1, characterized in that Before determining the similarity value between the feature vector of the target media content and the feature vector of the candidate media content, the method further includes: Determine a feature vector of the target media content based on the multimodal information of the target media content; wherein the multimodal information of the target media content includes information of multiple dimensions extracted from the target media content; And, based on the multimodal information of the candidate media content, determining the feature vector of the candidate media content; wherein the multimodal information of the candidate media content includes information of multiple dimensions extracted from the candidate media content.

3. The method according to claim 1, characterized in that Before determining the similarity value between the feature vector of the target media content and the feature vector of the candidate media content, the method further includes: Based on the identification of the target media content, a feature vector of the target media content is obtained from a target database; wherein the feature vector of the target media content is determined in advance based on the multimodal information of the target media content, and the target database also stores a feature vector determined based on the multimodal information of the candidate media content; Accordingly, determining the similarity value between the feature vector of the target media content and the feature vector of the candidate media content includes: Determine similarity values ​​between the feature vectors of the target media content and the feature vectors of the candidate media content in the target database.

4. The method according to claim 2 or 3, characterized in that: The multimodal information includes at least two of the following: a media content title, a media content text, a media content picture, and a media content video frame.

5. The method according to claim 1, characterized in that The determining the recommended media content based on the similarity values ​​corresponding to the candidate media content includes: Adding the candidate media content whose similarity value is not less than a preset first threshold into a target data pool; wherein the target data pool at least includes candidate media content determined based on target user data, and the target user data includes user interaction data and / or user interest tags; Determine similarity values ​​between feature vectors of the target media content and feature vectors of candidate media content in the target data pool; The candidate media content in the target data pool whose similarity value is not less than a preset second threshold is determined as the recommended media content.

6. The method according to claim 5, characterized in that Before determining the candidate media content in the target data pool whose similarity value is not less than a preset second threshold as the recommended media content, the method further includes: Determine the matching degree values ​​between the target user data and the candidate media content in the target data pool respectively; Correspondingly, determining the candidate media content in the target data pool whose similarity value is not less than a preset second threshold as the recommended media content includes: Determining a fusion score value corresponding to the candidate media content based on the similarity value and the matching degree value corresponding to the candidate media content in the target data pool; The candidate media content in the target data pool whose fusion score value is not less than a preset third threshold is determined as the recommended media content.

7. The method according to claim 1, characterized in that The candidate media content belongs to any of the following content themes: article theme, video theme, and graphic theme.

8. A media content recommendation device, characterized in that: The device comprises: A receiving module, configured to receive a media content recommendation request; wherein the media content recommendation request carries an identifier of a target media content, the target media content is determined in response to a trigger operation for a target notification message, and the target notification message is used to notify the target media content; A first determination module, configured to determine a similarity value between a feature vector of the target media content and a feature vector of a candidate media content, and determine a recommended media content based on the similarity value corresponding to the candidate media content; The returning module is configured to return the recommended media content in response to the media content recommendation request.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device implements the method according to any one of claims 1 to 7.

10. A media content recommendation device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

11. A computer program product, characterized in that The computer program product comprises a computer program / instruction, and when the computer program / instruction is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Multimedia recommendation method and device and electronic equipment

    CN115618024A

  • Template recommendation method and apparatus, device, and storage medium

    WO2022268133A1